Modeling And Simulating Of Mineral Grade Based On Floatation Froth Image Processing

The mineral resource development of our country has some problems such as low rate of resource utilization, laggard technology and equipment etc. The core of developing the production process inspection and automation of mining industry is researching high-ranking measuring instruments. Due to complexity, multi-parameters and multi-variables in mineral process, mineral grade is a crucial parameter for production guidance, economizing medicament, controlling product quality and raising recovery etc.With the development of computer technology, digital image technology and image collection equipment in recent years, the operator's operation through observing artificially floatation froth has not met the computer control requirement in flotation process. Computer vision replacing persons'eyes and utilizing digital image technology to promote mineral process becomes a new developed tendency and plays an important role on intelligent control in mineral plant.This thesis researches mainly the establishment for image processing system of floatation froth and modeling and simulating on mineral grade by characteristic parameters. The method of modeling on mineral grade uses support vector machine (SVM) to solve the nonlinear problem which can't be solved by the method of linear regression and also rises apparently to the forecast precision and ability of mineral grade and makes it adopt to the change of the industrial condition.